Distribution: Gentoo Linux Hardware Environment: NETGEAR FA-311, uses the natsemi 10/100MBps module, Intel's 10/100/1000MBps (onboard). Software Environement: gcc (GCC) 3.3.2 20040119 (Gentoo Linux 3.3.2-r7, propolice-3.3-7) - same with gcc-3.4 20040204), 2.3.3_pre20040207 using NPTL instead of LinuxThreads, linux-2.6.2-rc2-mm1. Problem description: trying to boot the kernel it stalls at initialization of eth0/eth1. While eth0 in case the both appropriate eth0-iface modules are built into the kernel image, and eth1 if they both are compiled as modules. I believe waiting for about an hour for it to load is enough to claim it stalls. Furthermore, the cursors stops blinking immediately. Steps to reproduce: try to boot a 2.6.2-vanilla or 2.6.2-mm1 or later using the .config I hope I'll be able to attach along with the init script for eth interfaces (/etc/init.d/net.eth0).
Created attachment 2075 [details] The used eth init script This same script is also used for eth1.
Created attachment 2076 [details] The used .config file In this .config the both eth drivers are to be compiled as modules, thus eth1 should be blocking the boot. eth1 is the Intel's gigabit eth interface device, eth0 is the NETGEAR's natsemi compatbile eth 10/100 megabit interface device - that is if they both are compiled into the kernel like it is now, since I can't see how they both are recognized by the not booting kernels.
Since my mozilla is not completely set up, I've jsut noticed I can't download the attachments properly, therefore I've placed the two files on my pub ftp: [1] http://neonik.net/gentoo/net.eth0 [2] http://neonik.net/gentoo/.config-2.6.3-rc1
Partitially resolved in this bug report: http://bugme.osdl.org/show_bug.cgi?id=2106
Does this still occur in 2.6.5-rc?
(mail exchange with NeoNik) >Does this still occur with say a 2.6.9 kernel? > This issue was solved as you can see there: http://bugme.osdl.org/show_bug.cgi?id=2106#c7 and on. So it does not occur any longer.